A graph-based semi-supervised learning for question-answering

Çelikyılmaz, Aslı; Thint, Marcus; Huang, Zhiheng

doi:10.3115/1690219.1690247

Cited by 22 publications

(24 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This component identifies and classifies the type of a given question among a set of predefined question types. The type of a question typically guides the search strategy at the phase of passage retrieval [2] or decides which types of named entities to extract at the phase of answer extraction [3]. To improve recall, one strategy Table 2: Features extracted from three phases of answer generation pipelined system: question analysis (S1), passage retrieval (S2), and answer extraction (S3).…”

Section: System Object Graphmentioning

confidence: 99%

See 1 more Smart Citation

A phased ranking model for question answering

Nyberg

2013

Proceedings of the 22nd ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

We describe a general result ranking approach for multiphase, multi-strategy information systems, which has been applied to the task of question answering (QA). Many information systems incorporate multiple steps and each step or phase may incorporate multiple component algorithms to achieve acceptable robustness and overall performance. Such systems may produce and rank a large number of candidate results. Prior work includes many models that rank a particular type of information object (e.g. a retrieved document, a factoid answer) using features specific to that information type, without attempting to make use of other non-local features (e.g. features of the upstream information source). We propose an approach that allows each phase in a system to leverage information propagated from preceding phases to inform the ranking decision. This is accomplished by a system object graph which represents all of the objects created during system execution, object dependencies (e.g. provenance), and ranking feature values extracted for a specific object. We evaluate the effectiveness of the proposed ranking approach in a multi-phase question answering system built by recombining pre-existing software modules. Experimental results show that our proposed approach significantly outperforms comparable answer ranking models.

show abstract

Section: System Object Graphmentioning

confidence: 99%

“…If we detect that they may represent the same concept or related concepts, their weights can be proportional to the similarity. Based on a result similarity function, we can replace w ki with Sim(R k , Ri) in Equation 3. In this paper, we define the similarity function as…”

Section: Relevance Votingmentioning

confidence: 99%

A phased ranking model for question answering

Nyberg

2013

Proceedings of the 22nd ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

show abstract

“…In order to formulate the semi-supervised truth discovery as an optimization problem, we choose the loss function based on studies on semi-supervised graph learning [18] [19][20], which have been widely used in many applications such as question answering [3] to image annotation [14].…”

Section: Problem Formulationmentioning

confidence: 99%

Semi-supervised truth discovery

Yin

Tan

2011

Proceedings of the 20th International Conference on World Wide Web

146

104

View full text Add to dashboard Cite

Accessing online information from various data sources has become a necessary part of our everyday life. Unfortunately such information is not always trustworthy, as different sources are of very different qualities and often provide inaccurate and conflicting information. Existing approaches attack this problem using unsupervised learning methods, and try to infer the confidence of the data value and trustworthiness of each source from each other by assuming values provided by more sources are more accurate. However, because false values can be widespread through copying among different sources and out-of-date data often overwhelm up-to-date data, such bootstrapping methods are often ineffective.In this paper we propose a semi-supervised approach that finds true values with the help of ground truth data. Such ground truth data, even in very small amount, can greatly help us identify trustworthy data sources. Unlike existing studies that only provide iterative algorithms, we derive the optimal solution to our problem and provide an iterative algorithm that converges to it. Experiments show our method achieves higher accuracy than existing approaches, and it can be applied on very huge data sets when implemented with MapReduce.

show abstract

“…Using textual entailment analysis, we obtain entailment scores between a natural language question posed by the user and the candidate sentences returned from search engine [1]. The textual deduction between two sentences is assessed via features representing high-level attributes of the entailment problem such as sentence structure matching, question-type named-entity matching based on a question-classifier, etc.…”

Section: Introductionmentioning

confidence: 99%

“…A SSL to demonstrate that utilization of more unlabeled data points can improve the answer-ranking task of QA. The graph for labeled and unlabeled data using match scores of textual entailment features as similarity weights between data points [1]. A summarization method applied on the graph to make the computations feasible on large datasets.…”

Section: Introductionmentioning

confidence: 99%